Jörg Hoheisel, Head of Division, DKFZ focuses on pancreatic cancer. Pancreatic cancer is one of the leading causes of cancer-related deaths in North America and Europe. Furthermore, the five-year survival rate typically ranges from 5 – 10%, therefore it is imperative that more research is done in this area to develop better treatments.
Hoheisel began by outlining the use of blood-based analysis for pancreatic cancer, including DNA methylation and liquid biopsy. He mentioned that his team achieved high accuracy in identifying cancer markers using machine learning. Some markers even reached an AUC value of 98% and a panel of six markers achieved an AUC of 1.0 which equates to a 100% accuracy value in tissue. This was harder to replicate in liquid biopsy. Yet, the team still managed to successfully translate this to circulating free DNA (cfDNA) in blood, with similarly high accuracy, but this was based on a limited sample size.
Next, Hoheisel discussed antibody profiling using protein arrays. They analysed around 3,500 antibodies covering around 2,500 proteins to identify tumour antigens and diagnose cancer. Although the profiling demonstrated high specificity, it displayed poor sensitivity.
Developing this work further, they wanted to further investigate whether this would help contribute to the decision on whether to operate or not operate on patients with IPMN, an early lesion of the pancreas that is considered a premalignant condition. So Hoheisel used a panel of 5 proteins and 3 microRNAs to improve diagnostic accuracy for IPMN from 80% to 97%.
Typically around 20% of patients with pancreatic cancer undergo surgery but reoccurrence of the tumour is fairly common. So, Hoheisel wanted to uncover whether it was possible to predict when reoccurrence may happen. To explore this, the team conducted protein profiling following surgery and showed that their methods can detect recurrence about 3.5 months earlier than current clinical practices.
Although some of these methods still require validation through clinical trials, the findings represent a significant step toward minimally invasive and timely interventions for a gruelling disease with historically poor outcomes.
